- 18 May, 2009 12:14 PM EST
- EA Stats For Fun
Our last poll may have seemed a bit arbitrary, asking you as it did to pick the most critical EA practice from a list of valid EA practices (i.e., "which EA element do you believe to be most critical?").
Out of 67 participants, here are the results:
Solid business vision: 51.1%
EA documented and communicated: 24.4%
Demonstrable financial impact: 17.8%
Completed gap analysis: 4.4%
EA measurement and reporting program: 2.2%
I asked this question because I wanted to see how it stacked up against some research I'm working on now. The selections available on the poll corresponded to questions from the enterprise architecture program maturity self assessment (G00150665). "Solid business vision" corresponds to question 17. "EA documented and communicated" corresponds to question 9. "Demonstrable financial impact" corresponds to question 27. "Completed gap analysis" corresponds to question 21 in the maturity assessment. EA measurement and reporting program corresponds to question 26. Granted this is work in process, and comparing a quick poll against data from 500 self-assessment records is a bit of a stretch, so take it with a grain of salt. It is, however, interesting.
The table below shows some of the data from the 500 records we've analyzed, along with some statistical information.
We are running various types of analysis against the data – including regression analysis. Looking at the data serendipitously, we can use regression weight as a way to think about the influence of a question and its response on overall maturity (i.e., a way to consider what matters most). One could conclude that gap analysis is most important because it has the highest regression weight, however, going to the actual questionnaire, most respondents are somewhere between reactively responding to specific requests for gaps and having identified some uncoordinated gaps.
Similarly, an EA measurement and reporting program seems like it should be high on the list as well. It has a fairly high regression weight. Looking a bit deeper, however, most selected the responses that suggested EA reporting was somewhere between ad hoc and exception driven to "some metrics delivered formally." Coupling this information with the lower regression weight of question 27 would lead one to wonder what is being reported formally or ad hoc if it is not financial efficiency metrics? The most probable explanation is that practitioners are getting reasonable traction by sharing success anecdotes or exposing problems in lieu of formal EA metrics programs. One could assert that reporting value, progress against goals, and successes is quite important.
Solid business vision, which clients ranked highest actually appears to be of least significance looking at our data. That said, the average response was 2.69, with a standard deviation of 1.2, we might as well say most people had a documented and agreed to business vision. At 2, business vision exists. At 3, it is documented and agreed to. At 4, it's documented, agreed to, and occasionally revisited and updated. So, because most have it, it probably doesn't influence the overall scores that much. It is table stakes.
I'll save the rest of the details for research notes. As Mark Twain said, "There’s three types of lies . . ."
